(1) Field of the Invention
The invention relates to a measurement system which quickly and nondestructively characterizes the mode-dependent losses and coupling of a multi-mode, graded index, connectorized, passive fiber optic component, and the modal sensitivity of a pigtailed optical source.
(2) Description of the Prior Art
Multimode fiber optic components are used primarily in low bit-rate, short haul communications systems. Current loss measurement techniques are unable to accurately predict the performance of a multimode component when utilized in a fiber optic system because such techniques are unable to account for mode dependent losses which often are exhibited by the multimode components.
A well known technique which at least partially characterizes a component's modal properties is the mode transition matrix method. The matrix formalism characterizes the modal properties of emitting components, such as light emitting diodes, using column vectors, detecting components, such as photodiodes, using row vectors, and passive components, such as connectors and couplers, using square matrices. The evolution of the optical power can be traced through an optical system composed of several components by multiplying each component's vector, or matrix, in reverse order of its physical sequence. For example, FIG. 1 illustrates a simple prior art optical system, including typical component matrix and vector values, which can be used to demonstrate the use of the mode transition matrix method. In the example, the emitting device is an 850 nm light emitting diode; the passive component is the coupled port of a 3 dB power splitter; and the detector is a silicon photodetector. The values of the vectors and matrix are obtained from measurements of the near field pattern as well as the total optical power both entering and exiting a passive component. The power propagation through the system is given by: ##EQU1## and therefore the loss is calculated to be: EQU loss=-10 log (P.sub.out /P.sub.in)
Although the mode transition matrix method improves the ability to predict system loss over conventional methods, the technique is inherently complicated. The near field pattern measurements can be accomplished using a charge coupled device array or a closed circuit television camera, and the data must be processed using complicated mathematical operations, such as curve-fitting, differentiation, and integration. Each time a different component is tested, new measurements of the near field pattern are needed. This technique of measurement, hereinafter referred to as the conventional mode block representation, is well-known and documented.